Describe an app in plain English. AI builds one that works.
See it The result
Live and real: built by this project's AI agent (Grok Bot) from the prompt above on 14 Oct 2026. On 15 Oct 2026, after an outside review, we replaced its first 'good enough' settle-up with an exact search for the fewest payments. Runnable checks: 1,808 pass. Not yet rebuilt in Claude/ChatGPT/Gemini by us; that consumer test is queued.
3 steps Try it yourself
Copy The prompt
Build me a simple web app I can open on my phone: a bill splitter for 5 friends after dinner.
- I can type the 5 names and how much each person actually paid.
- I can add the items we ordered with prices, and tick who shared each item (default: everyone).
- I can add tip % and tax %, split proportionally to what each person ate.
- Show what each person owes and the fewest payments needed to settle up ("Anna pays Ben €12.30").
- Handle cents correctly so the totals add up exactly.
- Everything stays on my phone: no login, no saving data online.
Make it a single page that works when I open it.Stuck? Where to paste, and what to expect
Where to paste or upload
On claude.ai the prompt box is filled in by our link; just press Enter. If you use another chatbot, paste the prompt into its message box.
What success looks like
A panel opens beside the chat showing a working bill splitter: names, prices, tick boxes, and a result that changes as you type.
If it goes wrong
No preview? Reply: “Please show this as an app I can run.” Numbers look off? Reply: “Show the calculation for each person.” Hit a usage limit? Wait for the window to reset or try another chatbot.
Heads up Limits & privacy
AI-built apps can have bugs (rounding, edge cases). Check a total by hand before trusting it with money. Don't type real bank details. Our demo is one currency only.
Privacy tips: what to remove from bills, CVs and photos first
Proof Test record
Verified: ✓ Built & tested by us
Live and real: built by this project's AI agent (Grok Bot) from the prompt above on 14 Oct 2026. On 15 Oct 2026, after an outside review, we replaced its first 'good enough' settle-up with an exact search for the fewest payments. Runnable checks: 1,808 pass. Not yet rebuilt in Claude/ChatGPT/Gemini by us; that consumer test is queued.
Read the full test record: exact prompt, inputs, raw output, failures and test code →
This didn't work for me
Tell us what happened. We'll re-test, fix the page, or mark it as broken, and note the date.